You will get expert Water Quality Index predictions with XGBoost & SHAP
Project details
I build ML pipelines to predict Water Quality Index (WQI) from your physicochemical data. Using XGBoost, Random Forest, and Linear Regression, I identify the best model (R² > 0.90) and explain key drivers via SHAP analysis.
DELIVERABLES:
• Custom WQI calculation (WHO standards)
• 3-model comparison with 5-fold CV
• SHAP explainability report
• 9 visualizations (distribution, spatial, trends)
• Interactive Folium pollution map
• Streamlit dashboard (cloud deployed)
IDEAL FOR: Environmental agencies, oil/gas companies, NGOs, municipalities monitoring river health.
REQUIREMENTS: Your water quality data (Excel/CSV with pH, DO, heavy metals, etc.) or coordinate boundaries for data collection.
TECH: Python | XGBoost | SHAP | Folium | Streamlit | MLflow
See my Niger Delta portfolio: 16 rivers, 35 stations, 93.9% accuracy.
DELIVERABLES:
• Custom WQI calculation (WHO standards)
• 3-model comparison with 5-fold CV
• SHAP explainability report
• 9 visualizations (distribution, spatial, trends)
• Interactive Folium pollution map
• Streamlit dashboard (cloud deployed)
IDEAL FOR: Environmental agencies, oil/gas companies, NGOs, municipalities monitoring river health.
REQUIREMENTS: Your water quality data (Excel/CSV with pH, DO, heavy metals, etc.) or coordinate boundaries for data collection.
TECH: Python | XGBoost | SHAP | Folium | Streamlit | MLflow
See my Niger Delta portfolio: 16 rivers, 35 stations, 93.9% accuracy.
Machine Learning Tools
MLflow, pandas, Python, scikit-learn, XGBoostWhat's included
| Service Tiers |
Starter
$400
|
Standard
$500
|
Advanced
$1,000
|
|---|---|---|---|
| Delivery Time | 2 days | 2 days | 3 days |
Number of Revisions | 1 | 2 | 3 |
Number of Model Variations | 1 | 3 | 3 |
Number of Scenarios | 1 | 2 | 4 |
Number of Graphs/Charts | 3 | 9 | 11 |
Model Validation/Testing | |||
Model Documentation | |||
Data Source Connectivity | |||
Source Code |
Optional add-ons
You can add these on the next page.
Fast Delivery
+$200 - $600
Additional Revision
+$150
Additional Model Variation
(+ 1 Day)
+$300
Additional Scenario
+$100
Additional Graph/Chart
+$75
Executive PDF Report
+$250
Live Training Call
+$400
3-Month Support
+$900Frequently asked questions
About Ebingiye Nelvin
Data Scientist | Python, SQL, Machine Learning, AI & Data Engineering
Port Harcourt, Nigeria - 7:59 pm local time
Core strengths:
Data analysis & data mining across large, messy, multi-source datasets
Machine learning & AI model development (classification, regression, predictive analytics)
SQL and Python-based data engineering ingestion, transformation, and clean data pipelines
API development & integration (REST APIs, third-party data sources, JSON/XML/CSV)
AI agent development, automation workflows, and LLM/RAG-based systems
Git/CI-CD workflows, Docker, and reproducible, production-ready code
Data visualization and dashboards that make findings usable for non-technical stakeholders
What I focus on:
I take ownership end-to-end from raw, inconsistent data through to a finished model, pipeline, or dashboard and I communicate tradeoffs and limitations clearly to both technical and non-technical stakeholders, so decisions are made on solid ground rather than a black box.
Proven work:
✔ Renewable Energy Site Selection System — built a machine learning model (91.25% accuracy) to identify optimal solar and wind farm locations, with a real-time dashboard across 400+ sites.
✔ Urban Heat Island Detection Pipeline — built a scalable data pipeline to process temperature/environmental data, delivered through an interactive dashboard for planning decisions.
I don't just hand over a model or a script, I help you understand what it means and how to use it.
Steps for completing your project
After purchasing the project, send requirements so Ebingiye Nelvin can start the project.
Delivery time starts when Ebingiye Nelvin receives requirements from you.
Ebingiye Nelvin works on your project following the steps below.
Revisions may occur after the delivery date.
Requirements & Data Intake
Client sends water quality dataset (Excel/CSV), station coordinates, and goals. I review data quality and confirm scope within 24 hours.
Data Cleaning & WQI Calculation
I clean your data, handle missing values, and calculate Water Quality Index using WHO standards. Share initial report for approval.